The impact of hybridization sequencing on the field of DNA sequence analysis, in general, and on the Human Genome Project, in particular, will be substantial. Once fully implemented, hybridization sequencing will result in a reduction in the cost-per-base by at least an order of magnitude compared with currently employed methods. Much of that savings will be realized through the elimination of materials and labor required for DNA sequencing using conventional gel-based approaches. In addition, further gains in throughput will be realized from the elimination of a number of limitations imposed by gel-based sequencing methods. As it is currently being developed, hybridization sequencing is flawed in several respects. First, it requires the synthesis of at least 65,536 interrogation probes, limiting the practical application of the technology to high-cost "DNA chip"-based products. Second, due to the size of the DNA fragments targeted for interrogation by hybridization sequencing methods, currently under development, control over what is known as the "branching" problem, is seriously impaired, giving rise to unacceptably high probabilities of failure. We propose an alternative approach in which a total of only 2048 probes are required to interrogate kilobase-size fragments with a very high probability of success. The method we propose is compatible with both "DNA chip", as well as hybridization membrane, technologies. Once developed, the hybridization sequencing product we propose will be extremely cost-effective and immediately applicable to even small, modestly-equipped laboratories.